These programs:
- Canadian Tech Growth Fund: $500M
- Compute Access Fund (SME): $700M
- LIFT (BDC financing): $500M
Will exist in the same way everything else does.
They will be rife with councils, middle men, gate keepers, and long procedures that will result in large companies who can navigate the bureaucracy benefiting from it, while startups, entrepreneurs, and innovators failing to access it.
We shouldn't be celebrating if @cohere gets funding from this.
We need to celebrate when pre-seed and early-stage companies that can generate revenue are being successfully subsidised due to innovating and taking risks.
We can't fix the country when only 5% of funds actually make their way back to the people that are establishing these markets instead of incumbents.
Really excellent work by the inference team to serve this model so efficiently!
To a significant degree, we have to become an AI inference company now.
Whether or not Europe stands with us, whether or not your journalists do their jobs, whether or not your politicians demonstrate the courage to act, I will fight for my people and my country.
Human-in-the-loop superintelligence arrived before AGI. Unlike predictions of @raykurzweil and @bengoertzel. Only problem is, this primitive form of it is super sloppy.
Communist governments like Iran's regime do a version of this. Put people on bullshit jobs and distribute the oil money. Not a uniform distribution obviously.
Universal HIGH INCOME via checks issued by the Federal government is the best way to deal with unemployment caused by AI.
AI/robotics will produce goods & services far in excess of the increase in the money supply, so there will not be inflation.
@0xSero A public dataset of steering data would be a huge win for us local LLM crowd. Labs have these internally as a result of their code subs. Would we be able to take a model that needs steering and add IQ (extra layers?) with such dataset?
@0xSero@FrameworkPuter Dope. I have a beelink Strix Halo, down for experiments. Most interesting thing imho is distributed training, e.g. with @tplr_ai. And ofc getting Pi run fast on long context. Here's where I'm at:
https://t.co/vtXYz3tmVM
FWIW, I'm really happy that a subset of the human population has arrived at a point where these are our concerns. Hope the rest catch up from barbaric fights over the dark liquid or whose god is right.